This is the authors' abstract. We don't add key points for this paper.
Surface Nitrogen dioxide (NO2) is strongly associated with various detrimental impacts on human health and the environment. Most current satellite-derived surface NO2 products predominantly utilize the vertical columnar density (VCD) of NO2 retrieved from low-earth orbiting hyperspectral satellites (e.g., OMI, TROPOMI, etc.) as the major predictor, and often experience limitations in terms of coarse temporal frequencies and spatial resolution. The capability of surface NO2 products to accurately assess the temporal and spatial variability of NO2 concentration is subsequently constrained. We proposed a novel conception to generate surface NO2 concentrations with high temporal (<= hourly) and spatial (250m) resolution from multispectral satellites (e.g., AHI and MODIS) using machine learning approach, leveraging the coexistence and chemical and/or physical interconnections among various pollutants. The approach was tested in the urban agglomeration regions of China characterized by severe pollution. The model was validated against surface measurements with the high and stable accuracies and outperforms most models developed in previous similar studies using VCD of NO2 measurements. The model generated NO2 product has been effectively employed to continuously monitor the temporal and spatial evolution of NO2 pollution during the air pollution episodes and locate the sources of NO2 emission. The proposed approach broadens the applications of MODIS and AHI satellites (and potentially others) in atmospheric environment research, providing a potential alternative pathway for the advancement of ground-level pollutant products through satellite remote sensing.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: